CHAU, Calvin, Jan KŘETÍNSKÝ and Stefanie MOHR. Syntactic vs Semantic Linear Abstraction and Refinement of Neural Networks. Online. In Automated Technology for Verification and Analysis. ATVA 2023. Singapore: Springer, 2023, p. 401-421. ISBN 978-3-031-45328-1. Available from: https://dx.doi.org/10.1007/978-3-031-45329-8_19.
Other formats:   BibTeX LaTeX RIS
Basic information
Original name Syntactic vs Semantic Linear Abstraction and Refinement of Neural Networks
Authors CHAU, Calvin, Jan KŘETÍNSKÝ (203 Czech Republic, belonging to the institution) and Stefanie MOHR.
Edition Singapore, Automated Technology for Verification and Analysis. ATVA 2023, p. 401-421, 21 pp. 2023.
Publisher Springer
Other information
Original language English
Type of outcome Proceedings paper
Field of Study 10201 Computer sciences, information science, bioinformatics
Country of publisher Germany
Confidentiality degree is not subject to a state or trade secret
Publication form electronic version available online
Impact factor Impact factor: 0.402 in 2005
RIV identification code RIV/00216224:14330/23:00133938
Organization unit Faculty of Informatics
ISBN 978-3-031-45328-1
ISSN 0302-9743
Doi http://dx.doi.org/10.1007/978-3-031-45329-8_19
Keywords in English Abstraction; Machine learning; Neural network
Tags core_B, firank_B
Tags International impact, Reviewed
Changed by Changed by: RNDr. Pavel Šmerk, Ph.D., učo 3880. Changed: 8/4/2024 10:14.
Abstract
Abstraction is a key verification technique to improve scalability. However, its use for neural networks is so far extremely limited. Previous approaches for abstracting classification networks replace several neurons with one of them that is similar enough. We can classify the similarity as defined either syntactically (using quantities on the connections between neurons) or semantically (on the activation values of neurons for various inputs). Unfortunately, the previous approaches only achieve moderate reductions, when implemented at all. In this work, we provide a more flexible framework, where a neuron can be replaced with a linear combination of other neurons, improving the reduction. We apply this approach both on syntactic and semantic abstractions, and implement and evaluate them experimentally. Further, we introduce a refinement method for our abstractions, allowing for finding a better balance between reduction and precision.
PrintDisplayed: 5/9/2024 01:23